The general advice that goes around is to look at things like $50 mixing decks and such. Well I have news for you, Corrosive Truths is a budget production. Now, when I say it’s a budget production, I don’t mean, well, we have to make do with only a couple of Shure SM58s, a two-input mixing deck, and a low-latency sound-card specifically for our podcast, no no. I mean proper, shoe-string, budget. I mean, I want to buy as little equipment as absolutely necessary whilst still being listenable. A secondary purpose was just to make sound guys cry. This is achieved easily enough by ignoring most of their advice for what it is, fidelity snobbery. Well, sound quality wasn’t our biggest concern, having a podcast with three to four people using only existing equipment and spending under about £20, was.
What was a king to do?
Well, much of the advice out there says things like; don’t use the microphone port on your computer, buy a better sound card with little to no latency, use these software products from Adobe, spend at least $50 on a mixing deck with a good brand.
We proceeded to; use a computer microphone solution, used an existing sound card, better yet, the on-board one that came with the computer, used Audacity (which is completely free), and spent exactly $0 on a mixing deck with a good brand.
Specifically, we use four “Mini Compact 3.5mm Jack Audio MIC Microphone for Skype MSN Google VoIP Windows XP/Vista/7 Webcam Vedio Internet Call Laptop PC with 1m One meters cable“. Lovely title, isn’t it? These we bought from Amazon for the princely sum of £11.92 for all four. What else? Well you can’t record right next to the back of the computer, and those microphones only have meter long cabling, so an extension cable was needed. This wasn’t the specific one we bought, but a newer, and marginally cheaper version at £3.49. Of course, we have four microphones and one extension cable, and one port. Now for the part that’ll make the sound guys weep once again, we use splitters to join the signals into one signal. Worse, two-way splitters were cheaper, so we bought three. That way you can have a single two-way split should you only need two microphones, otherwise, plug two microphones each into two splitters and those two splitters into the third. We bought the splitters from a local electronics outlet, but you can get something very similar, again from Amazon at £0.67 per unit. If we add up all those prices, we come to a grand total of … £17.42.
The sound quality we can get out of this equipment after some post production, whilst not being the best, is still quite good.
I should note that the equipment recommendations of most websites, whilst not being really true to the spirit of doing things on a budget, are generally sound advice for those of you who want good end results. For example, Peter Reviews does use a Shure SM58 with a Lightsnake cable (XLR to USB, so simple) for the higher fidelity. Good cheap alternatives are SingStar and Rock Band style microphones.
Before we start, we’ll do volume checking, making sure each microphone is giving us about the right volume, if someone is quiet, we’ll simply swap the microphone with the loudest person. It doesn’t have to be exactly equal because we use lots of dynamic compression to get the voices sounding equal (it’s only voice, so it’s fine to be heavy with the plugin). We keep recording short clips, adjusting, then compressing until it’s right.
Generally, we have a selection of topics to work with and we’ll, as a group, look at each topic and work out what the pertinent questions are and then just start talking about them. All the better if there’s an obvious flow from one question to another, but we don’t concern ourselves too much about sticking rigidly to any schedule. We prefer to keep the recording going pretty much of the time, but will occasionally stop if we want to have a longer off-mic discussion.
Editing is very light and quick, longer pauses are removed, stuff that shouldn’t be on is taken out, noise reduce if needed (generally the built-in sound reduction from the sound card itself is good enough). After that, it’s dynamic compression, again, pretty much the defaults, but with the rate turned up to full. We use ffmpeg to compress the file at 96k mp3, encoding the version that goes onto youtube is a bit more awkward:
ffmpeg -i audio.wav -loop_input -i picture.jpg -s hd720 -ab 128k -g 300 -qscale 2 Ep002.mp4
I won’t go into lots of detail, this loops a single picture, resized to 720 for time t, which we get from using ffmpeg -i on the audio file. With the g option we are telling ffmpeg to encode in such a way that static imagery has better compression, so we can get away with using less bits for video, there is no bitrate, we just want it to keep good quality, this is the qscale option. At that point, the files are all ready for upload.